How to constructively integrate speculative ideas in your portfolio.
After making several thousand percent on your money in cryptocurrencies, the one thought on your mind is: „How can I do that again?“
Wild feelings are running amok, but recent signs of a bitcoin top may suggest looking elsewhere for crazily volatile speculation.
Bitcoin daily price January 21st 2018 having broken through 9 and 60 day moving averages and posting lower lows and lower highs
Or just maybe you weren’t so lucky (prescient?) and have only been participating in crypto for the last month and are down thousands of dollars? The cryptocurrency rally is quite unique in the way that hardly anyone participated in it – at least before that last leg above $10.000. I don´t know anyone who did myself.
Being a realist, making another thousand percent in Bitcoin from today´s price seems unlikely to you and the argument that the cryptocurrency market looks a tad bubblicious starts to makes sense.
The Uranium Opportunity
Before watching the rest of those profits go down the drain, isn´t the obvious choice to take all your money and put it into Uranium mining stocks for another virtually certain 10x return? Where else in the world is there such a deep value opportunity left?
Even left-for-dead coal mining ETF KOL had it´s 200% run starting in 2016…
Here is the setup:
- The one sign for great value in a commodity: production costs are far higher than the price of Uranium
- Major producers cutting production reducing supply: Cameco and Kazhakstan
- Newly built global nuclear power capacities are set to cause increased demand
- A bombastic breakout into a new uptrend from record low prices in Uranium ETF URA
- A retracement giving a great risk reward trade with a stop loss below all-time lows or the long term moving average
So BUTT (Back up the Truck) as is the conclusion in this hilarious and enlightening post by the macro tourist!
By the way, the virtually exact same setup last year was an utter failure…
But, does it still make sense to take such a trade – apart from the obvious ethical and environmental considerations?
Or is there a risk-controlled way to participate in Bitcoin just in case it does run up again and repeats 2017´s crazy move?
Position Sizing
Maybe a more rational position size than risking everything would be appropriate? Legendary investor Bill Miller, for example, advocates putting a mere 1% of your investment capital into bitcoin. At such a small allocation it won´t matter much even if it goes to zero.
On the other hand how much influence will that position have on your portfolio? Well, hardly any – if your volatile speculation does go up 1000% in a year (which is incredibly rare) your portfolio grows by 10%. That´s hardly the life changing wealth we expect from such a return, rather it´s pretty much the average annual gain from a boring old index fund.
If you judge uranium ETF (URA) as less likely to go to zero as long as nuclear energy is in use (the big risk factor is a new Fukushima type event) you could risk a higher position, let´s say 5%. The return expectation could be based on what a similarly beaten down energy commodity returned recently (100% per year for coal KOL), this would mean a 5% return for the portfolio as a whole.
Again, not something to make us jump into the air with joy and go quit our job.
Risk-Adjusted Return
Investor attention is drawn to highly volatile, extraordinary returns. Often the expected return for speculating on such an asset is overestimated – investors mistakenly extrapolate recent performance into the future. But such superior performance has a relatively small probability of actually materializing as most of the gains have been made before the idea became visible to the mainstream. As much as this over-expectation is true for momentum speculations (“look how much this has gone up, let´s jump on the train and not miss any more of these gains”), it also is for mean reversion (“this thing has gone down so much, it must be the greatest opportunity ever”).
What really matters is the expected risk-adjusted return for each position: the expected return for your speculation divided by its volatility will give an expected Sharpe ratio (the most commonly used measure for risk-adjusted return) for each investment which makes an apples to apples comparison to other investments and even whole strategies possible.
An Unrealistic Example
The hard part is reaching a realistic return expectation for a high conviction speculation, because it just sounds so convincing. Let´s use a highly optimistic guess, that could still possibly come true, as an example:
- Bitcoin: expected return 200%, recent volatility (1 year) 46%, expected Sharpe ratio 4,35 (200/46=4,35)
- Uranium ETF URA: expected return 100%, recent volatility 23%, expected Sharpe ratio 4,35 (100/23=4,35)
- S&P 500: expected return 20%, recent volatility 8%, expected Sharpe ratio 2,5 (20/8=2,5)
Now we can at least judge that hitting a home run in a speculation has better prospects than a great performance of a stock index – the expected risk-adjusted return is the same for both speculations: almost double the S&P´s. The big question that remains, the answer to which is the essence of investing, is: which event is more probable?
Even if this is very uncertain, we can still use the information we gathered (asset volatility) combined with our world view (expected return) to construct a portfolio with some diversification and better position sizing. Let´s assume the probabilities, that we actually reach our return expectations, are equal and returns from our different investments uncorrelated.
We can now use the individual investor´s version of the modern portfolio theory to create an efficient portfolio: a smart rule of thumb.
First we can weight positions by expected risk-adjusted return (the higher it is, the more we want to invest). The speculations each get an equal weighting and the S&P only half as much as we expect it to return about half on a risk-adjusted basis: 40% each for Bitcoin and URA and 20% for the S&P.
But as each asset has a vastly different volatility, that we need to take into account. We want to get an equal volatility weighting, so that each position has the same influence on the return of our mini-portfolio. Because URA has half the expected return and half the volatility of Bitcoin, it should have double the actual position size as Bitcoin has for the same risk and return impact.
The S&P has only about 1/6th the volatility of Bitcoin, but only gets to have half the impact on the portfolio, because it´s expected risk-adjusted return is half. Putting these two numbers together we can calculate an actual position size of three times Bitcoin´s. That gives us a simple recipe for the position sizes making up a portfolio that includes our different return expectations and volatilities for each asset (rounded percentages):
- 1 part Bitcoin 17% (100/6 as 6 parts are 100%)
- 2 parts URA 33% (100/(2×6))
- 3 parts S&P 50% (100/(3×6))
By quantifying our expectations we can tame the urge to pour all our money into the vastly higher returning Bitcoin (which might turn out to be a huge loss as the asset is so volatile) and end up with half our money in the S&P because it is the safer, least volatile bet, even though we expect it to return less. Our overall allocation has a much higher chance of success because it is diversified with each position contributing the same amount of risk to the portfolio – each position going spectacularly wrong will have the same impact.
Even if this is quite a weird portfolio, we have started to integrate our speculative ideas into a defined process – we moved away from thinking about individual positions towards creating a balanced portfolio. We would get the same result if we were only half as optimistic in our return expectations and we could easily integrate a judgment about which investment is more likely to be successful, if we have an opinion about that. We can now apply these ideas to a healthier portfolio solution.
Repeatable Process and Sound Principles
Good investing and trading is not about one-off great outcomes, but rather a repeatable process. Thinking things through from a portfolio perspective is a huge step into the right direction.
There is nothing wrong with overweighting high conviction plays to satisfy our inner speculator`s irresistible urge, but it´s a good idea to keep it from running rampant. Putting in a foundation of sound investment principles to integrate speculative ideas gives them a much greater chance to be successful.
Outside View on Chances of Speculative Ideas
In our hearts we know how unlikely it is to make a life changing return from a one-off investment. But very often the optimistic conviction that this time it´s different, that this particular investment has a much higher chance of superior, even life changing, performance, clouds our judgement. One of the guiding principles of my investing process, inspired by Daniel Kahnemann and Michael Mauboussin, is looking at the world outside to try to judge the general basic probabilities – in this case how high the success rates of speculative ideas are in general. Speculative browsing led me to an interesting data source on high conviction picks by investment experts who are making investment recommendations in newsletters. They will usually tout their newest ideas with the highest possible upside in their sales pitches, as they know that greed and FOMO lead to a high number of subscriptions. They will also try to pick an idea that they are really convinced of, because if it actually delivers as promised they will likely keep their readers and gain recommendations.
Travis Johnson runs an interesting website called Stock Gumshoe where he very successfully analyzes these newsletter teasers to find out the names that are being touted. He also compiles these into a database which allows you to track the performance of these pitches over the last 10 years.
I looked at high conviction pitches as a proxy for an individual investor´s speculative ideas and analyzed their success rate:
The overall return over 1 to 3 years after the pitch was very close to the S&P 500 with a surprisingly similar number of winners and losers (60% winners and 40% losers with 13% of recommendations reaching a 100% or higher return). It is interesting to observe that older recommendations massively underperformed after 2-3 years.
My conclusion is that speculative ideas have a good, but not extraordinarily high chance to win big (around 10% to 15%, compared to about 1% of the S&P companies returning 100% in 2017), but also a high probability to suffer big losses – returning in-line with the index on average. They are not well suited for long term investment – after the hype has run its course many winners fall in price.
Return Factors
Academic studies about the characteristics of different outperforming return factors provide a sound base to investing. Let´s look at three of them in the context of our speculation example.
Whenever I look at my portfolio steady, boring positions usually top the performance ranking – only very rarely do I see exciting speculations there, even though they seem to draw a lot of my attention. At the moment, for example, top performer is timber ETF WOOD, which is part of my commodity allocation. Historically (1990 – 2009) timber as an asset class has shown the highest Sharpe ratio (0,96) of all asset classes and the ETF (investing in timber companies, not the commodity itself) does keep chugging along quietly.
Maybe my speculative ideas are especially bad, but studies support this empirical observation:
1 Low Beta Factor: Low volatility assets outperform high volatility assets.
This means that there is a basic bias against speculative ideas which are high beta in nature. We should be smart enough to underweight them even if we are very optimistic about them, but this might be a bit of a battle against the inner speculator.
2 Value Factor: A pick like URA based on favorable valuation has a sound principle working in its favor. Statistics compiled by Meb Faber say that it is extraordinarily rare for an asset to be down more than 5 years in a row as Uranium has been. His basic rule of thumb is that, if an asset is down five years in a row, you can expect future two year total returns of around 100% on average.
3 Momentum and Trend: Bitcoin´s strong surge also points to a probable superior future performance as outperforming assets continue to outperform on average.
A Better Possibility to Integrate Speculation in a Portfolio
As shown in the unrealistic example above, a good portfolio process that deliberately weights each individual position can be used to integrate all ideas holistically – even extreme speculation.
In detail this is a complex process and you can find practical applications throughout this blog. I also recommend Robert Carver´s book on the subject “Smart Portfolios: A practical guide to building and maintaining intelligent investment portfolios.“
Let´s stick with our example and look at how I might integrate these ideas in different parts of my portfolio to give them the right size and impact – the principles can be applied in any portfolio construction process, even if the structure is totally different.
At the top level of my portfolio I tactically allocate across three major investment and trading strategies – at the moment these are equally weighted:
Global Asset Allocation GAA – Short Volatility – Trend-following Managed Futures
I need to find a place for each idea in one of these buckets and scale the position size accordingly.
URA and WOOD could be part of the real assets in the GAA (1/3rd of my portfolio) – real assets are weighted at 1/3rd of GAA at the moment, altogether 11% of capital is invested in real assets. Within these 11% I can use expected Sharpe ratios and volatilities of all real assets to determine my position size for these two ETF. To make things easier, let´s say 5% of my portfolio are invested in other real assets already leaving 6% to allocate to the two ETF.
- WOOD has a volatility of 10% and a long term Sharpe ratio of 0,96 which we will expect for the future.
- URA, as above, is optimistically expected to return 100% with recent volatility at 23% giving us an expected Sharpe ratio 4,35
This leads to a rough allocation of 2% WOOD and 4% URA in the portfolio (URA 4x better Sharpe ratio but twice as volatile = 2x WOOD) – I use generously rounded numbers as guesstimates are a big part of the equation anyway. Because WOOD has shown its merits over the long-term and URA´s expectations are biased upwards by a good story, I might be smart and adjust to equal weight. I could even switch the percentages around listening to my experience more than to hope, but as they say: hope dies last.
In any case the impact on the portfolio will be so small that I can´t really make a profoundly bad decision.
Bitcoin is much harder to evaluate simply because it´s such a new, rapidly developing asset, but as it recently became part of the futures market I can easily integrate it into my trend-following managed futures strategy. The strategy has a very clear allocation scheme which uses the average true range over the last 60 days to calculate a volatility weighted position size for each asset. I don´t need to have any return expectations and am only concerned with the risk impact the position has on my portfolio – if Bitcoin goes up again, profits will take care of themselves. For the exact strategy rules please refer to this post. As soon as I get an entry signal (when the 9 day moving average (MA) re-crosses over the 60MA) I go long (calculation per $350.000 of investment capital as the contract is quite large and volatile):
0,33% x capital / ATR(60) x contract multiplier = 0,33% x 350000 / 1230 (ATR from Yahoo Finance BTC-USD) x 1 = 1 contract long
This position should move 0,3% of the portfolio value on an average day, while risking the loss of approximately the same amount initially.
In comparison URA´s risk impact should be about 0,1% per day on average (on January 24th 2018 4% of $240.000 buys 660 shares @ $14,50; with ATR (60)= 0,334 this translates into an average daily move of $220), which sounds about right as the GAA strategy is more long-term in nature with a much wider stop loss than the Trend-following strategy uses.
This shows two concrete ways to integrate speculative ideas into a complex overall portfolio and avoids that we get carried away by over-optimism about a fresh idea.
Conclusion
You don´t have to follow every new thing, whether it´s a deep-value-commodity-fad or a life-changing-new-technology-fad.
There are opportunities everywhere and boring steady assets and strategies will likely be the outperformers. But if you do need to chase speculations now and again it´s best to do it systematically within a sound portfolio construction process.
One comment